The General Theory of Employment, Interest and Money

The book’s central quantity is the return an investor expects from newly built capital. Keynes defines it against two things and only two: the yield the asset is expected to produce over its life, and what it costs to build one today. The rate is forward-looking by construction. What a comparable asset actually returned in the past does not enter the definition, which is a sharper restriction than it first appears — it means a track record is evidence about the future, never a substitute for an estimate of it.

This expected return falls as more of a given type of asset gets built, and Keynes separates the two reasons. The expected yield declines because output from the new capacity competes with output from the rest. The cost of building declines more slowly or rises, because the facilities that produce that type of capital come under strain. Over a short horizon the cost pressure dominates. Over a longer one, the falling yield does.

The passage most worth carrying out of the chapter concerns what an expectation of cheaper money later does to the value of building now. Keynes argues that expecting rates to fall in future lowers the expected return on equipment built today, because that equipment will spend part of its working life competing against equipment built later and financed more cheaply — machinery that can accept a lower return and still clear. Anyone financing hardware with a short useful life against a long-dated obligation is inside this sentence.

The link to interest rates is stated plainly and is the book’s operational claim: new investment is pushed to the point where the expected return on the marginal asset has fallen to the rate of interest. Two variables, one equality. Most of what follows is an argument about which of the two moves more.

His answer is that the expected return is the unstable one. The rate of interest is a market price, observable and comparatively steady. The expected return is a forecast held by people who cannot know what they are forecasting — and it is capable of collapsing quickly, for reasons the next chapter takes up.

Expectations of yield rest partly on facts that can be known — the existing stock of capital, current demand — and partly on future events that can only be guessed at with varying confidence. Keynes is explicit that the second category dominates precisely where the sums are largest and the assets longest-lived.

In practice investors do not solve this problem; they adopt a convention. The convention is to assume the present state of affairs continues indefinitely, except where there is specific reason to expect a change. Keynes’s point is not that this is stupid — it is a reasonable response to genuine ignorance — but that the stability it produces is borrowed. It holds exactly as long as the convention holds.

Organised markets revalue holdings continuously, which changes the character of a commitment. Keynes compares it to a farmer who could tap the barometer after breakfast and withdraw his capital from farming between ten and eleven, then reconsider later in the week. An investment that is fixed for the economy as a whole becomes, for any individual holder, something he can leave.

He draws the distinction the rest of the chapter runs on. Enterprise is forecasting an asset’s yield across its whole life. Speculation is forecasting what other people will think. He observes that as investment markets get better organised, the risk that the second crowds out the first goes up rather than down.

The mechanism is not irrationality but incentive. A professional who believes an asset’s prospective yield justifies 30 will still not pay 25 for it if he expects the market to mark it at 20 in three months. Forecasting the convention is the paying activity; forecasting the yield is not. Keynes treats this as the predictable result of how the market is built, not a failure of character.

Hence his attack on what he calls “the fetish of liquidity” — the doctrine that institutions should hold readily saleable securities. His objection is arithmetic before it is moral: there is no such thing as liquidity for the community as a whole. Every holder can sell only if some other holder buys. The individual’s exit is real; the aggregate’s is not.

He is candid that the fetish has a use. Because each holder believes his own position is liquid, he is willing to take risks he would refuse if the commitment were permanent. Remove the belief and new investment would be harder to raise. Keynes calls this a dilemma and does not resolve it.

Where no solid basis for calculation exists, he expects sentiment to move in waves of optimism and pessimism that are “unreasoning and yet in a sense legitimate”. That last clause is the useful one. It denies the analyst the comfort of calling a swing simply a mistake.

Keynes splits the two sides of the credit market. The schedule of expected returns on new capital governs the terms on which funds are demanded for investment. The rate of interest governs the terms on which funds are supplied. They are separate mechanisms that meet at a price, and confusing them is the error he spends the next chapter on.

The rate of interest, in his account, is the price of parting with liquidity rather than the reward for saving. That reframing is the book’s most contested move and its most consequential: it makes the rate a monetary phenomenon set in the market for money, not a real one set by thrift.

He gives three motives for holding cash rather than lending it. The transactions motive covers ordinary business needs. The precautionary motive covers the wish for certainty about the future cash value of part of one’s resources. The speculative motive covers the attempt to profit from judging the future better than the market does.

The third is the volatile one, and it produces a dilemma parallel to the one in Chapter 12. Without an organised market for debt, precautionary demand for cash would be far larger. With one, the speculative motive gets room to swing hard. Organising the market does not remove the instability so much as relocate it.

Keynes argues that deriving the rate of interest from the expected return on capital is circular. In equilibrium the two are equal, since investment expands until they are. But the expected return depends on the scale of current investment, and the scale of current investment cannot be worked out without already knowing the rate. He notes Marshall ran into the same wall.

What survives the demolition is the equality itself, used in the right direction: given a rate, investment proceeds until the marginal expected return has fallen to meet it. The equality is a consequence to reason forward from, not a theory of where the rate comes from.

Keynes points out that every durable commodity has its own implied rate of interest — wheat, houses, anything storable — and that the money-rate has no special claim on paper. The chapter asks why, then, output and employment should be tied so tightly to the money-rate in particular.

The answer turns on money’s peculiar properties: it cannot readily be produced by hiring labour, and demand for it does not fall away as its value rises. Those two features let the money-rate act as a floor the other rates must clear, which is why it ends up setting the hurdle for real investment.

Keynes locates the crisis in a sudden collapse in the expected return on capital rather than in a rise in the rate of interest. The rate usually rises too, but he treats that as an accompaniment. The break is in the forecast, not the discount rate applied to it.

The collapse is violent because the optimism it replaces was never resting on much. Expectations built on a convention rather than a calculation can fall a long way quickly, and the same absence of a solid basis that let them rise lets them drop.

He is unimpressed by the remedy of raising rates to prevent the boom, on the grounds that it treats a problem of misjudged returns with an instrument aimed at the cost of funds. The recovery, in his account, has to wait for the expected return to be rebuilt, which takes as long as it takes.

The book closes on the rate of interest as a policy variable rather than a natural constant. The traditional case for a moderately high rate was that saving must be induced; Keynes has argued that the amount actually saved is determined by the scale of investment, which a low rate promotes. The justification therefore falls away.

The qualification matters and is often dropped when the chapter is quoted: he attaches the recommendation to a limit, the point beyond which stimulating investment this way stops corresponding to anything real.

Chapter 11’s argument — that expecting cheaper money later reduces the value of building now — is the cleanest available statement of the problem facing anyone funding short-lived hardware with long-dated debt. It was written about industrial equipment in 1936 and it describes a data centre without adjustment.

Chapter 12 supplies the vocabulary for a market that funds a buildout eagerly at the index level while individual lenders step back from individual deals. Enterprise and speculation are asking different questions, and Keynes’s claim is that a well-organised market systematically rewards the second.

The book pairs with Big Debt Crises, which tracks what borrowing does to a balance sheet, and with Irrational Exuberance, which tracks how the story justifying it spreads. Keynes supplies the middle term: why the expected return underneath both is the least stable number in the calculation.

Applied on this site at The Credit Wall.